gh@utai.UUCP (Graeme Hirst) (09/28/84)
AI Seminar, Department of Computer Science, University of Toronto
REASONING ABOUT KNOWLEDGE AND BELIEF
Kurt Konolige
Artificial Intelligence Center
SRI International
Menlo Park, Calif. 94025
Tuesday 2 October 1984, 3pm, Sandford Fleming 1105
Reasoning about the knowledge and beliefs of computer and
human agents is assuming increasing importance in Artificial
Intelligence systems for natural language understanding,
planning, and knowledge representation. A natural model of
belief for robot agents is the deduction model: an agent is
represented as having an initial set of beliefs about the
world in some internal language and a deduction process for
deriving some (but not necessarily all) logical consequences
of these beliefs. Because the deduction model is an expli-
citly computational model, it is possible to take into
account limitations of an agent's resources when reasoning.
We investigate a Gentzen-type formalization of the
deductive model of belief. Several original results are
proven. Among these are soundness and completeness theorems
for a deductive belief logic; a correspondence result that
relates our deduction model to competing possible-worlds
models; and a modal analog to Herbrand's Theorem for the
belief logic. Specialized techniques for automatic deduc-
tion based on resolution are developed using this theorem.
Several other topics of knowledge and belief are
explored from the viewpoint of the deduction model, includ-
ing a theory of introspection about self-beliefs, and a
theory of circumscriptive ignorance, in which facts an agent
doesn't know are formalized by limiting or circumscribing
the information available to him.
(I will give an overview of the deduction model, and then
proceed to any of the sub-topics above that are of
interest.)
--
\\\\ Graeme Hirst University of Toronto Computer Science Department
//// utcsrgv!utai!gh / gh.toronto@csnet-relay / 416-978-8747